From International Socialism, 2:4, Spring 1979.
'Here we have the old constitutional folly. The condition of a "free government" is not the division but the UNITY of power. The machinery of government cannot be too simple. It is always the craft of knaves to make it complicated and mysterious.' – Karl Marx 
For some years now, there has been a welcome revival in Marxist discussion on the State. Two names, above all, have been associated with that discussion: Ralph Miliband and Nicos Poulantzas. 
Criticism of the work of both authors has been abundant. Among the more perceptive critics  it was noted, separately for each writer, that his work was open to reformist interpretation. Both produced works from which revolutionary socialists could learn, but the way in which each argued his case was such that reformist lessons could not be ruled out. This possibility did not arise because of the readiness of reformists to distort anything they come across, and pervert it. (Though that is a well-known phenomenon.) Rather, the very way in which Miliband and Poulantzas constructed their accounts of the state and of capitalist society was itself ambiguous from this viewpoint.
Rather schematically, we can see where the heart of the problem lay. Poulantzas in one of his own contributions to the New Left Review, suggested that a key problem with Miliband was his lack of any 'theoretical problematic'. He was, in fact, quite wrong: Miliband's The State in Capitalist Society revealed that he had a very distinct 'theoretical problematic', a 'theory of society'. That theory was quite different from Marx's, however. Poulantzas failed to notice this, above all, because he made the same fundamental mistake as Miliband – though on different grounds.
What was the mistake? At root, it was a failure to comprehend Marx's enormous theoretical breakthrough, in particular by comparison with the classical political economists to whose critique he devoted a large part of his theoretical life work. Marx's achievement in this sphere lay, above all, in his ability to get behind the categories of political economy – value, capital, property, rent, state, class, etc. – and to show that these expressed historically created social relationships. In particular, Marx showed that 'production' could only be adequately understood if it were seen as a social process through which human beings create and recreate their own world. Far from being simply a 'technical' relationship between human beings and nature, production was also centrally a social activity through which 'men' – in the sexist language of the 19th century – made their own history, their own society.
To understand production, it was necessary to grasp its various social forms. Thus, at the beginning of Capital, Marx analyses the commodity as the expression of a double production process: it is, simultaneously, a use-value, a thing of use to human beings, and a value, the expression of a certain kind of social relationship among producers. In producing commodities, we produce not only useful things, but also a distinct form of society founded on alienated social relations.
Similarly, all the other language of political economy has to be reconstrued to reveal this active, social creative process. Workers in capitalist factories make cars, steel, chemicals, toothbrushes, etc.: but in this very act of making things, they also produce surplus value. Their work activity takes a form in which they reproduce their bosses, producing the means by which their own exploitation and domination is continued. The whole social order – relationships of family, state, science, education, etc. – should be understood as perpetually produced and reproduced elements made by real active individuals in their social interconnections. It is not that one section of society makes society, makes the environment, etc., while another is merely the passive, organised section. All human history is the record of activity of all the individuals who compose it.
Hence, for Marx, revolutionary socialist politics is grounded in an understanding of history. The very possibility of socialism – of a society in which the whole population makes its own society consciously and in accordance with a democratically determined plan – rests on the analysis of capitalist and other class societies: for what men and women have made and remade they can remake again in line with their own self-developed needs and perceptions. However radical a break with the past a socialist revolution may be, it does not introduce into history an entirely new principle, the self-activity of human beings. Without that self-activity, the whole of past history cannot be understood.
Thus Marx's more 'philosophical' writings are all of a piece with his revolutionary politics. His critique of Hegel, and of mechanical materialism, is all of a piece with his critiques of political economy and of various 'socialisms from above'.
Within this perspective, it is the particular form that social relationships take which defines the various 'modes of production'. In class societies, like capitalist society, the key to comprehending society is the form that the active struggle between the classes takes. In particular, the modes of action and the social relationships among the oppressed majority are crucial to comprehending how the society is maintained, and how it may be overthrown through revolutionary practice.
Now, while the analyses of Miliband and Poulantzas are not the same, it is apparent that in neither of these authors' work is this central understanding of the class struggle, and in particular of the forms of activity of the exploited classes, present as a central and defining element. Thus there is a measurable distance between their analyses and those of Marx.
In Miliband's case, The State in Capitalist Society rests on a theory of society which – as one of his more acute critics, Isaac Balbus  pointed out – combines 'elite' theory and 'stratification' theory. That is, the whole study is organised around the theme, not of class struggle, but class domination. In Miliband's study, only the ruling class appears to act. In the various institutions which he analyses, there is little or no sense of a conflict of classes. (Given that his book was published in 1969, for instance, it is remarkable that his account of education in schools and colleges says nothing about student revolt…) Thus his study is very one-dimensional in its emphases. When he discusses the bases of existence of the ruling classes, he emphasises not their role in production but their benefits from distribution. Thus the ruling classes are primarily defined in terms of their (passive) possession of riches rather than their (active) role as capitalists. The working class, similarly, is defined as the class that 'works hardest and gets least', not as a class whose struggles and forms of organisation shape the form of society. The working class as an active, creative, struggling class does not appear in the body of his analysis: there it merely suffers. So, when it suddenly pops out on the last page of the book, as the force which will one day wipe out capitalism, the idea appears as a decorative addition rather than the conclusion of a single argument.
As for the state, Miliband devotes little space to a discussion of the various institutional means by which the present state excludes the working class from power. Rather, he emphasises its class character by the methods of orthodox sociology: by examining not its forms but the 'social class' origins of its upper members, their possession of attitudes akin to those of the rich, etc. Thus, as John Lea  pointed out, he leaves open the question whether the 'machinery' of the state can be captured and used by working class parties for the benefit of the working class and to destroy capitalism. He leaves unexplained and uncriticised the very forms of the state – its characteristic bureaucracy, the 'democratic sham', the nation-state form, etc.
The work of Poulantzas, on the other hand, rests on a system of thinking developed by the 'Althusserian' school, a system which has already been the subject of a number of important critiques.  The Althusserian system represented, in one sense, a reaction against an interpretation of Marxism which it called 'economism' – the view that the historical process is the product of changes in the 'forces of production', i.e. a form of determinism. It is characteristic of 'economism' (or what Colletti termed 'the Marxism of the Second International' ) that production is treated as a technical and not simultaneously a social process. However, as Simon Clarke suggests (see note 3) the attempt to avoid the trap of 'economism' is not necessarily to fall on the ground of Marxism. There are other possible traps. Indeed, Althusser leaps into another hole, one in which even the works of Marx himself have to undergo a major process of rewriting and rereading if they are to be purged of their innumerable failings.
In place of 'economism', Althusser proposed the idea that societies be considered as complex systems of interdependent levels – the economic, the political and the ideological – all of which mutually interpenetrate each other but over which the economic is determinant 'in the last instance' – which 'last instance', Althusser assured his readers, never comes. It has been pointed out by endless critics that Althusser's system strongly parallels the system developed by the conservative American sociologist, Talcott Parsons, that it is a 'structural functionalism'. It certainly is a system, as Parsons' is, in which it is difficult to locate principles of historical change. History, as a 'process without a subject', consists of a series of modes of production each of which is treated as an 'eternity'. Nothing within each mode threatens it: to believe so is to fall into the trap of 'economism' or 'historicism'. The system is radically elitist in its emphases: Althusser makes a distinction between 'science' and 'ideology' according to which all forms of society require ideology, so that socialism too will have to be a system which is ultimately impenetrable to all but its few 'scientific' members.
Poulantzas took over this system. His central account of the state was cast in functionalist terms: it is 'the global factor of social cohesion'. The state, as both repressive and ideological apparatuses, holds the whole system together. Not only this, but the state actually forms and structures the relations of production, forming isolated individuals and thus the relations of competition which characterise the capitalist form of society. It IS not the case that the state in any sense 'rises out of' or 'expresses' the social relation's of capitalist production and exchange: it constitutes these relations. In taking over the Althusserian system, Poulantzas added something which in Althusser is chiefly decoration: classes. But his classes are formed in the 'relatively autonomous' "political-juridical level", not within the relations of production. In the 'economic level' there occurs a production process – the production of goods, based on technically conceived 'relations of production' consisting of a combination of workers, instruments and objects of production, and non-workers. Nowhere in Poulantzas's analysis do we find an account of the class struggle understood as a set of social relations through which people produce capitalist society. Social relations are politically constituted, in the 'relative autonomy' of the political-juridical sphere.
Despite the considerable differences between Miliband and Poulantzas (which differences extended to their comprehensibility), they shared a common starting point. Both began with technically conceived relations of production, onto which they then constructed social (class) relationships of distribution, of ownership and appropriation. As a result, both of them had difficulty in explaining change within their analytical frameworks. Thus both failed to relate their discussion of fascism (a brief section in Miliband, a whole book by Poulantzas) to the inter-war crisis of world capitalism; neither examined the forms of the class struggle under Hitler or Mussolini. Both, despite considerable arguments between them, offered similar analyses of ideology and 'ideological apparatuses', which they both presented as internally coherent and non-contradictory, and certainly not as ideas shaped and reshaped in the process of struggle between the classes.
To borrow a term from Althusser, there is a very significant 'absence' in both of them. Neither gives an account of the class struggle, rooted in the social production relations of capitalism, as the key organising and disorganising set of social relations in capitalist society. In neither's analysis, therefore, does the actual struggle of workers play any significant role. In that crucial sense, the analytical frameworks of both writers were always open to reformist political interpretation.
Until recently, however, final judgement on them had to be held. But both writers have now taken the political plunge quite directly, publishing books with explicitly reformist political programmes. Miliband concludes his recent Marxism and Politics with some reflections on the Chilean debacle. The lesson he draws from the downfall of the Allende regime is, in essence, that the same experience must be repeated; only, next time there will have to be present a grouping of socialists who will point out to the reformist government the necessity of deepening the process of social transformation in its own defence. Thus a social revolution will be achieved without all the Sturm und Drang of an actual revolution. 
And now Nicos Poulantzas has published a book, State, Power, Socialism  whose appearance announces that the famous "Miliband-Poulantzas debate" is, in all essential respects, over. For the two of them are in basic political agreement with each other. Both have placed themselves on the "left wing" – insofar as that is a meaningful term – of the present "Eurocommunist" tendency. Both have announced, unambiguously, their breach with the Marxism that asserts, with Marx, that 'the emancipation of the working class must be conquered by the working class itself.'
Poulantzas suggests that the transition to socialism must proceed on two levels. On one hand, the parliamentary system must be both used by the Left, and maintained as an integral part of socialist politics. On the other hand, and in parallel, there must be a development of workers' councils or 'self-management' bodies, organised on the principle of direct democracy. The problem is to find 'a democratic road to socialism, a democratic socialism',
posed as follows:
how is it possible radically to transform the State in such a manner that the extension and deepening of political freedoms and the institutions of representative democracy (which were also a conquest of the popular masses) are combined with the unfurling of forms of direct democracy and the mushrooming of self-management bodies 
The 'fundamental problem' is that of 'combining a transformed representative democracy with direct, rank-and-file democracy'.  'Democratic socialism is the only kind possible.'  At the national level, central government is to be organised on a parliamentary basis. This, he makes clear in the Interview with Henri Weber, involves a government elected by universal suffrage, on a secret ballot, with no 'imperative mandate' over MPs, no right of recall over MPs , rights for bourgeois parties.  In addition to this national system, there are to be local bodies based on the principles of 'direct democracy', i.e. with recall of delegates, mandation of delegates, etc. Parliamentary, representative democracy nationally, plus workers' councils based on factories, etc: these two forms are to co-exist, and be articulated together, in ways Poulantzas admits  he can't at this stage specify very precisely:
the answer to such questions does not yet exist – not even as a model theoretically guaranteed in some holy text or other.
As he explains, this idea involves quite definitely abandoning the idea of smashing the existing state and replacing it with a state of the kind celebrated by Marx in the Paris Commune:
... the expression 'sweeping transformation of the state apparatus in the democratic road to socialism' suggests that there is no longer a place for what has traditionally been called smashing or destroying that apparatus. The fact remains, however, that the term smashing, which Marx too used for indicative purposes, came in the end to designate a very precise historical phenomenon: namely, the eradication of any kind of representative democracy or 'formal' liberties in favour purely of direct, rank-and-file democracy and so-called real liberties. It is necessary to take sides… talk of smashing or destroying the state apparatus can be no more than a mere verbal trick. What is involved, through all the various transformations, is a real permanence and continuity of the institutions of representative democracy – not as unfortunate relics to be tolerated for as long as necessary, but as an essential condition of democratic socialism. 
No more talk, either, of the 'dictatorship of the proletariat': Marx, who used the term, used it
as a notion of applied strategy, serving at most as a signpost. It referred to the class nature of the State and to the necessity of its transformation in the transition to socialism and the process of the withering away of the State. 
Continued use of the term 'dictatorship of the proletariat' would only obscure Poulantzas's programme: he therefore supports the PCF's[the French Communist Party] decision to drop the idea.
Therefore, Poulantzas criticises those in the PCF who want to retain the phrase in the party programme (e.g. Etienne Balibar). Such people, he suggests, are given to uttering 'dogmatic
banalities' of the following kind: '...every State is a class State; all political domination is a species of class dictatorship; the capitalist State is a State of the bourgeoisie.'  No doubt it is true that these 'dogmatic banalities' have, on occasion, been uttered without examination, with criticism, without deepening of their meaning. But does this make them incorrect? (The better part of Poulantzas's own work has been devoted to elucidating their truth, after all.) Poulantzas continues his discussion of these 'banalities' by suggesting that 'such an analysis is incapable of advancing research by a single inch', and that it can't help us understand 'concrete situations, since it cannot account for the differential forms and historical transformations of the capitalist State.'  This kind of simplification led to the disasters of Stalinism in the face of fascism, he suggests.
It has to be noted that Poulantzas refers to only one kind of 'disaster… in the face of fascism': the 'social fascism' analysis which underpinned the German Communist Party's appalling tactics in the face of Hitler. But can the 'social fascism' doctrine be attributed to the 'banalities' outlined above? Not without political mediations, it can't. Trotsky, for one, never doubted the truth of these 'dogmatic banalities', yet his analysis of the rise of Hitler, and his proposals for a united front to defeat the Nazis, have yet to be bettered, Furthermore, it was the 'forgetting' of these, 'dogmatic banalities' which led the Comintern in the Popular Front period (on which Poulantzas is quite silent) to grovel before the Blum government in France and to follow a strategy in Spain from which all the gains were made by General Franco. Poulantzas, however, can hardly criticise the Popular Front: his proposals are for a re-run of that experience.
Poulantzas's proposals, I want to suggest, would if pursued produce disaster. They would involve the working class repeating a whole series of defeats already experienced this century. His programmatic utterances announce his own, unambiguous, breach with Marxism. And his argumentation for them is very weak.
Does workers' power necessitate Stalinism?
Poulantzas suggests that the classical Marxist conception of socialist revolution – destruction of the existing state apparatus and its replacement by direct workers' democracy – leads directly to 'statism', i.e. to Stalinism. All the horrors of Stalinism, to which the communist parties are now becoming sensitive, were rooted in the way the Russian Revolution was conducted. This, of course, is a proposition that finds ready assent from a chorus of liberals and other anti-socialists, so Poulantzas evokes the seemingly impressive authority of Rosa Luxemburg. In 1918, this great German revolutionary socialist wrote a pamphlet strongly criticising the Bolsheviks for, among other things, their dissolution of the Constituent Assembly.
Poulantzas, however, does not offer much in the way of concrete data on the question. And a few matters have to be faced directly. Was Luxemburg's criticism of the Bolsheviks correct? Why did the Bolsheviks, who'd been the most active campaigners for the Constituent Assembly, then dissolve it by force? Were they wrong to do so?
On the first issue, that of Luxemburg's criticisms, Poulantzas fails to mention a couple of germane points. The first is that, according to some historians , she later retracted this specific criticism of the Bolsheviks. But it should also be noted that within a year of the Russian Revolution she was to be faced with exactly the same issue in Germany. There Kautsky's USPD made proposals of exactly the same kind Poulantzas now advocates: for a National Assembly plus soviets. Rosa Luxemburg was utterly uncompromising in her opposition to the idea. Poulantzas does not quote her opinion on the matter – not surprisingly.
Whoever pleads for a National Assembly is consciously or unconsciously depressing the revolution to the historical level of a bourgeois revolution: he is a camouflaged agent of the bourgeoisie or an unconscious agent of the petty bourgeoisie. 
The path of the revolution follows clearly from its ends. Its methods follow from its tasks. All power in the hands of the working masses, in the hands of the workers' and soldiers' councils. This is the guiding principle … Every act, every step must like a compass point in this direction. 
The National Assembly, rather than a guarantee of 'democratic socialism', was that 'counter-revolutionary stronghold erected against the working class.' 
Thus, in her own practice, Rosa Luxemburg did not believe in the maintenance of parliamentary forces. Had she suddenly lost her (1917) commitment to democracy? Far from it: her position arose directly from her profound commitment to democracy, from her belief that nothing should stand between the working class and direct power. Clearly, if we are to play the game of citing 'authorities', Luxemburg is a poor one for Poulantzas to quote. Why did the Bolsheviks dissolve the Constituent Assembly? They certainly had been campaigning for a Constituent Assembly to be convened during 1917. Indeed, their argument had been that only a powerful soviet system could guarantee a Constituent Assembly. It does not seem to have occurred to them, prior to October, that there might be a conflict between soviets and Assembly. In itself that isn't very surprising: the Bolsheviks were in the process of recasting Marxist theory in action. After all, they entered the revolution believing it would be a bourgeois revolution: before February, only the isolated non-Bolshevik, Trotsky, with his theory of permanent revolution, thought otherwise.
The October uprising completely altered the picture. Lenin was among the first to grasp this, urging a delay in the Constituent Assembly elections and extension of the vote to 18 year olds. As so often, he failed to win a majority for his proposal, and the elections went ahead. Overall, the Socialist Revolutionaries won a clear majority, though the strength of their vote was directly proportional to the voters' distance from the centres of the revolution. The SR victory was, at best, ambiguously democratic: in the uprising, the SRs had split, with the Left SRs supporting the uprising and the Right opposed, yet the elections lists were presented by a 'united' party. Thus it wasn't clear what the electorate had actually voted for.
On the Assembly's first day, the Bolsheviks pronounced ratification of the soviet seizure of power, the decree on land, the decree on peace, workers' control of production, etc. By 237 to 136 votes, the Assembly rejected the motion. The same day it was dispersed. It had become the rallying centre for counter-revolution, as it continued to be in the Civil War that broke out within months. Not to have dispersed the Constituent Assembly would have been a crime against the revolution: it would have permitted the re-emergence of a situation of 'dual power', with two rival national political centres, each with opposed aims and methods. 
To suggest, as does Poulantzas (who also repeats the old canard about Lenin's What Is To Be Done prefiguring Stalinism ) that it was this event which led to Stalinism is a most fantastic historical judgement and a piece of utter formalism. The tragedy of the degeneration of the Russian Revolution arose, above all, from the working class's loss of its ability to govern directly, under the impact of the isolation of the Revolution and the terrible rigours of the Civil War, the de-urbanisation, famine, collapse of production, etc. The Soviets, the living heart of the revolution, ceased to beat with the life of the Russian workers. Not, let it be emphasised, initially through the bureaucratic manipulations of any anti-democratic party or leadership, but under the terrifying pressure of circumstances. 
Poulantzas's second line of argument is his insistence that 'representative democracy', i.e. parliamentarism, is the key guarantee of the preservation of political freedoms. Thus it is the strategic line of defence against state authoritarianism. In his interview, he cites the opinion of the Italian social democrat Nobert Bobbio:
[Bobbio] did highlight one point. He said: 'If we want to maintain liberties, the plurality of expression, etc., then all I know is that throughout history these liberties have been coupled with a form of parliament.' Certainly he expressed it in a social-democratic form. But yet, I wonder if there isn't a core of truth in that, if the maintenance of formal political liberties doesn't require the maintenance of the institutional forms of power of representative democracy. Obviously they would be transformed: it's not a matter of keeping the bourgeois parliament as it is, etc. 
Poulantzas doesn't offer very much by way of arguments for 'representative democracy' beyond the observation that both this form of government and other political freedoms were 'conquests of the masses'. Historically, though we might cavil at this simplification of the path by which workers won civil and political rights under capitalism, we needn't object in principle. Marxists continue to defend those rights against attacks from the right. On the other hand, we also do not glamorise the extent or nature of these rights and freedoms. In particular, we have to recognise the very ambiguous and limited character of the 'conquest' involved in representative democracy. (Indeed, both in his earlier work, Political Power and Social Classes, and indeed in his latest offering, Poulantzas is not uninteresting on the whole issue: he provides plenty of materials for denouncing parliamentary government as sham democracy.) Let's recall a few of the major limitations on what is properly called 'bourgeois democracy'.
Workers in the capitalist democracies have the right to vote in parliamentary and local elections. This right they exercise through the secret ballot.  Thus each voter exercises his 'power' in isolation from any community, as an individual in an atomised relation to the state. This atomistic relation between 'citizen' and state was a major theme in Marx's critique of the capitalist state, from the early 1840s, and is indeed continued in Poulantzas's own writings.  Individual voting is not a matter for public discussion, or for meetings. It is utterly individualised. Revolutionaries have always argued that mass meetings are more democratic than secret ballots, since they permit the exercise of wider kinds of political reasoning than can be applied by the isolated voter. In a mass meeting, issues can be discussed, arguments refuted. In a mass meeting, estimates can be made of the general level of support for some proposals for action, and thus of the likelihood of that action being successful. This is not possible with the secret ballot.
Secondly, and notoriously, voters have no real control over their elected 'representatives'. There is no effective right of recall, no effective mandate which voters can exercise over MPs, etc. The voters elect the MP as isolated voters, and as such have no control over him or her. The MP is protected from control by constituents by a whole gamut of 'privileges' once he or she is elected. The MP ceases to represent anyone once elected. In the late 18th century, the conservative Edmund Burke expressed the relation between MP and constituents very well when he told the electorate of Bristol that if they voted him into Parliament he would not represent them "to the nation", he would represent the nation to them. More recently, Harold Wilson reportedly expressed the same idea with a definition of democracy as 'government of the people, for the people, by the people – with the emphasis on government'. Thus, in the parliamentary system, the exercise of political freedom and power consists in the few seconds, every few years, it takes the voter to express a choice between parliamentary misrepresentatives – marking his ballot paper with a cross, like an illiterate.
Thirdly, it is in any case only the legislature which is elected by the anonymous and powerless parliamentary constituencies. We exercise no control over the remaining part of the state machinery: if we gain some influence there, it is not by legal-constitutional means, but by means varying from riot to bribery. We do not elect our army, judiciary, police or humorously named 'civil servants'. Nor have we effective means of control over them. If complaints against them are made, most often they have their own internal mechanisms for providing their own judges and juries in their own cases. It is even exceptional for the legislature itself, i.e. Parliament, to exercise real control over the state bureaucracies. The non-elected part of the state – that vast machinery which we support with taxes and other exactions, and which maintains a multiplicity of controls over our daily lives – is protected from popular control by a whole variety of institutional means, including rules of 'contempt' (note that there is no charge of 'contempt of people' that can be brought against our judges), official secrecy (extending even to the rules for Social Security benefits), bureaucratic appointment, etc. Even the MPs are excluded from scrutiny of large parts of the bureaucracy's personnel and actions.
Fourthly, the legislature deals chiefly with the framing of laws, with general rules rather than particular cases. (Otherwise, as radio listeners have now discovered in Britain, it chiefly generates noise like a school debating society on an off afternoon.) But the business of modern states is more and more concerned, not with generalities and widely applicable laws, but with specifics, with the detail of bargains between ministries and corporations, with particular administrative cases and so forth. Parliaments are neither empowered nor competent to deal with these issues. The whole framework of the 'legal state' is thus being progressively undermined by tendencies to concentrate capital and power.  Orthodox political science has registered this development with a growing literature on 'pressure group theory', the 'decline of parliament', etc. Parliament, which never very precisely represented anyone, comes less and less to. represent real powers in society and to involve itself in real decisions.
In short, the 'conquest', of the vote has proved rather insubstantial. It cannot be said that the defence of the existing right to vote is a major defence of the democratic principle. Now Poulantzas, of course, wants to preserve a 'transformed' representative democracy, though lie is generally vague on how it is to be 'transformed'. What is not to be transformed, however, he insists, is the secret ballot, the lack of a power of recall over MPs, the lack of a mandate over MPs, etc.  So, wherever we are to look for a 'transformation', it is not in the direction of democratisation.
But then he claims he has an altogether grander aim in view. The task is 'to open up a global perspective of the withering away of the state'.  The idea of the 'withering away of the state' is a noble idea first developed by Marx and Engels, along with their other ideas about smashing the existing state, the dictatorship of the proletariat, and the other things Poulantzas wants to junk. It refers in Marx and Engels to the process whereby the overwhelming majority, the working class, having formed a new state power under their direct control, is then and only then capable of developing a new pattern of social relations and control over the conditions of life such that they stand in less and less need of organised violence to manage their social affairs. With the withering away of the state, conviction rather than coercion becomes society's key organising principle. But the maintenance of a parliamentary body, which is not even directly subordinated to society, does not even begin to meet the conditions for the realisation of this aspiration. It stands in the way, as an impediment to democratic life, as a governmental form out of the control of the people. In reality Poulantzas is not proposing a route to the 'withering away of the state': to suggest that he is, is, in his words, a 'verbal trick'.
In short, Poulantzas's 'democratic road to socialism' depends on the maintenance of an undemocratic form of government. 
Attack on workers' self-government
Poulantzas's argument also embodies an attack on direct democracy, on workers' councils as a form of government. Here too his arguments are very weak.
His first point is simply a non-argument: namely, that up to now all such movements of the working class, aiming at the dictatorship of the proletariat, have failed. It is of course true: there is no socialism anywhere in the world. By various means, every revolutionary movement of the working class has so far been beaten back. That is no more of an argument against learning from the historical experience of these defeats in order to succeed next time than drownings are an argument against learning to swim.
Secondly, he suggests that the existing state is too strong to permit the emergence of a rival centre in a 'dual power' situation, as a prelude to socialist revolution:
...if you consider the essence of the state apparatus as it is in France, and then the forms of centralisation of popular power… Well it's obvious that it will be crushed before it's taken more than three jumps of a flea. You surely don't think that in the present situation they will let you centralise parallel powers to the state aiming to create a counter-power. Things would be settled before there were even a shadow of a suspicion of such an organisation. 
Here Poulantzas introduces a whole new dimension into the dialectic: the method of self-contradiction, or having your cake and eating it. In support of his own argument, the state appears as essentially weak , too weak to be certain of barring the 'democratic road'. For his socialist opponents, however, the state is too strong!
The argument is anyway ridiculous. Who, after all, thinks that 'in the present situation' in France (or anywhere else) workers are going to try to centralise the power of their workers' councils? The very precondition of such a development is that the 'present situation' has changed. The idea of revolution in a non-revolutionary situation is absurd. Every revolutionary situation has involved a split within the existing state apparatus and the existing ruling class. A revolutionary situation involves a crisis for the state, a loss of effectiveness. Without such a crisis there can be no revolution: that is part of the ABC of Marxism. It is precisely the crisis in the state which permits the emergence of a situation of 'dual power' and the possibility of a new form of state power conquering.
In practice, Poulantzas – like all reformists – seeks to construe a transition to socialism from the 'present situation'. He suggests that today's capitalist state is best understood as a state in crisis, a state of crisis:
This state crisis offers the Left new objective possibilities of a democratic transition to socialism. There are several kinds of political crisis: the present one defines for the Left a precise field related to the possibility of a democratic transition. What is involved is neither a dual-power crisis nor a crisis stemming from a tendency towards fascism. 
There are two elements here: first, a correct (if 'banal') observation that general tendencies to crisis are not in themselves revolutionary crises; second the absurd idea that there can be a transition to socialism without a revolutionary crisis. By revolutionary crisis I mean not some economic slump, but a crisis in class relations of the type Lenin referred to in 'Left-Wing Communism': a boiling situation where the ruling class cannot keep ruling in the old way and the oppressed refuse to continue being ruled in the old way. Poulantzas sometimes recognises this. He suggests the election of a 'Left' government can only amount to a 'social democratic experience' unless there is simultaneously a mobilisation of the 'popular masses'. That this would fundamentally alter the 'present situation' for the working class, for the state, for all forms of political life, however, seems to escape him.
Thirdly, Poulantzas suggests – in line with his argument about Russia – that direct workers' democracy leads to Stalinism, the suppression of political liberties and the crushing of dissent:
Direct democracy, by which I mean direct democracy in the soviet sense only, has always and everywhere been accompanied by the suppression of the plurality of parties, and then the suppression of political and formal liberties. 
This theme is repeated in the book. If we 'base everything on direct, rank-and-file democracy' we take a path that 'sooner or later, inevitably leads to statist despotism or the dictatorship of experts' ; if workers' councils form their own state power, 'it is not the withering of the State or the triumph of direct democracy that eventually emerges, but a new type of authoritarian dictatorship'.  It must be said that Poulantzas offers no arguments in support of this elitist proposition that working class self-government is impossible, only flat assertions. But we can note two things.
First, Poulantzas is wrong. There has not been an association between workers' councils and the disappearance of a 'plurality of parties' (China, Cuba, Cambodia, etc., are nothing to do with the case, for there never were workers' councils there). The Paris Commune had a plurality of parties; so did the Spanish workers' councils in 1936-7; a key demand of the Hungarian workers' councils in 1956 was a plurality of parties.
Second, what Poulantzas is actually arguing for is a plurality of bourgeois parties. These, for him, are the guarantee of 'political liberties'. It is true there has not tended to be much room for bourgeois parties in workers' councils – though, not so much because workers' councils have banned them as because in revolutionary situations bourgeois parties do not seek democratic rights within workers' councils, but seek to destroy them! The Constituent Assembly, after all, refused to recognise the soviet power, and Mrs Thatcher's chief interest, faced with a congress of workers' councils in Britain, will hardly be one of winning a delegation from the Finchley working class …
It should also, perhaps, be noticed that Poulantzas doesn't even seem to understand the idea of a socialist revolution. At several points he employs a 'fortress' analogy which he attributes to the revolutionary left: e.g. 'It is first of all necessary to take state power, and then, after the fortress has been captured, to raze to the ground of the entire state apparatus, replacing it by the second power (soviets) ...' ; '...first of all the existing state power is taken, and then another is put in its place. This view can no longer be accepted.'  Whoever did 'accept' this notion? It is a reformist fantasy. The idea that a socialist revolution first puts itself in charge of the bourgeois state apparatus, then abolishes it is nonsensical. The existing state power is a target only for destruction, not for 'taking'.
But, does Poulantzas want socialism anyway?
It is by no means clear that Nicos Poulantzas, judging by his own arguments, doesn't want to avoid socialism altogether. In the Interview, he successively makes two points: first, and unexceptionably,
I agree with you: the whole of the present state and all its apparatuses – social security, health, education, administration, etc. – correspond by their very structure to the power of the bourgeoisie. I do not believe that the masses can hold positions of autonomous power – even subordinate ones – within the capitalist state. They act as means of resistance, elements of corrosion, accentuating the internal contradictions of the state.
and then the remarkable statement:
[It is necessary to struggle within the state] not simply in the sense of a struggle enclosed within the physical confines of the state, but a struggle situated all the same on the strategic terrain of the state. A struggle, in other words, whose aim is not to substitute the workers' state for the bourgeois state through a series of reforms designed to take over one bourgeois state apparatus after another, and thus conquer power, but a struggle which is, if you like, a struggle of resistance, a struggle designed to sharpen the internal contradictions of the state, to carry out a deep-seated transformation of the state. 
In short, no taking of power!
In his book, he proposes that the state's economic apparatuses should not be smashed:
At no point should changes lead to the actual dismantling of the economic apparatus: such a development would paralyse it and accordingly increase the chances of boycott on the part of the bourgeoisie. 
How awful, we might paralyse the Department of Trade and Industry, or the Treasury! Poulantzas feels able to argue this, despite the fact that on the previous page he has stated that the 'economic apparatus' (yes, the same one) 'remains, in its unity, an essential factor for the reproduction of capital'. Note, too, that the bourgeoisie is presumed still to have the power of boycott. Might not this risk have been reduced by the local workers' councils taking control of the factories, banks, etc.? No, this too is ruled out:
... the democratic road to socialism refers to a long process. the first phase of which involves a challenge to the hegemony of monopoly capital. but not headlong subversion of the core of the relations of production. 
The reader who dares to ask the naive question, "Why on earth not?" gets an answer: Change cannot go beyond certain limits without running the risk of economic collapse. 
Thus, the transition to socialism is to occur from the 'present situation' and without 'economic collapse'. And Poulantzas thinks revolutionaries are utopians! Against Poulantzas, we must be clear: a transition to socialism, to the complete reorganisation of society by the working class, cannot occur without 'economic collapse'. A socialist revolution involves 'economic collapse': the problem is to carry it through, decisively, so that economic recovery on a new basis can be started immediately.  But not for Poulantzas:
Over and above the breaks involved in the anti-monopoly phase, the State will still have to ensure the workings of the economy-an economy which will remain to a certain degree capitalist for a long time to come. 
Lessons of history
Parliamentary democracy and workers' councils, I suggested, are incompatible institutions. Poulantzas suggests they must be yoked together, though he doesn't know how. But, he remarks that 'History… has provided – and that is not insignificant – some negative examples to avoid and some mistakes upon which to reflect.'  Let's reflect, then – on a classic 'negative example' and the 'mistakes' involved in a 'combination' of parliamentary democracy and workers' councils: Spain in 1936 and 1937.
In July 1936 Franco's revolt against the Spanish republic began. Initially, the revolt was defeated over large parts of Spain, largely by popular forces who demanded (and often seized) arms from the Republican government. The background to the generals' revolt had been a development, all across Spain, of convulsive mass struggles. The signal for these had been provided by the election, in February 1936, of a Popular Front government. There had been a wave of general strikes and land expropriations, workers' and peasants' councils were formed, etc.
After July, this movement developed tenfold. In large parts of Spain, especially Catalonia, Aragon and Castile, workers' and peasants' councils organised production and distribution, ran the towns and villages, set up their own militia forces, etc. At the centre of republican Spain was an enormously weakened bourgeois parliamentary government. Spain, after July, represented a classic situation of 'dual power'.
The republic's chief source of military aid against Franco was Stalin. He wanted a diplomatic and military alliance with the ruling classes of France and Britain against Hitler, and was implacably opposed to the development of socialist revolution in Spain – out of fear, among other things, that this would alienate the French and British governments. In backing the Republic, therefore – and at a high price – he insisted that the revolution be contained within bourgeois democratic limits.
The period up to mid-1937 was one of permanent struggle between the central republican government, chiefly armed and organised through Moscow, and the local workers' councils, workers' militia, etc. The bourgeois democratic Republic was clearly incompatible with the workers' and peasants' struggles and organisations, and with the demands they embodied (socialised property, land expropriation, workers' control of production, etc.). The outcome of the struggle was that first the forces at the base of Spanish society were limited and contained, and then the workers' councils were destroyed in pitched battles, police actions, etc.: in Barcelona in early May 1937, in Aragon a little later on.
What the whole tragedy demonstrated was the clear incompatibility between the maintenance of the bourgeois parliamentary government and even local workers' councils, even local independent working class action that went beyond the narrow prescribed limits set by the central government. The key to the tragedy of Spain was that those who led the workers' councils – above all the anarchists – opposed the centralisation of the workers' councils, thereby leaving a void at the centre of Spanish life which the counter-revolutionary liberals and Communist Party filled, organised and used against the working class. From that struggle the only victor was General Franco.
Wherever the issue has appeared, parliamentary democracy has proven incompatible with forms of workers' councils. If one form is to be maintained, the other has to be destroyed. In France, in 1936, to save the Popular Front government, the Communist Party called off a mass strike movement. At the end of the war, in France and Italy, the Communist Parties saved the parliamentary system by disarming the Resistance. In Chile, the maintenance of the Allende Popular Unity government was achieved through attacks on the workers' movement, both directly and through the method of bureaucratising and limiting it. The results, over and over again, are famous disasters.
Poulantzas is not putting forward a 'new' strategy, but an old, tried and tested, fully guaranteed formula – for working-class defeat. It is the job of Marxists to insist that a mass workers' movement that does not complete the process of socialist revolution by centralising its own power in new institutions, and by smashing all obstacles to that centralisation of its power in workers' councils, prepares its own downfall. It is vital that the ideas of those like Poulantzas, who propose limitations on workers' power, be combated as strongly as possible.
It may not, however, be Poulantzas himself we have to combat. He ends his book with a gloomy reverie on the 'risks' of the 'democratic road':
… at worst, we could be heading for camps and massacres as appointed victim. But to that I reply: if we weigh up the risks, that is in any case preferable to massacring other people only to end up ourselves beneath the blade of a Committee of Public Safety or some Dictator of the Proletariat… There is only one sure way of avoiding the risks of democratic socialism, and that is to keep quiet and march ahead under the tutelage and the rod of advanced liberal democracy. But that is another story.
The ambiguity is characteristic. Is this to be the topic of Poulantzas's next offering, we wonder? It would not be inappropriate: for he has shown, decisively, that he is already under the theoretical tutelage of liberal democracy – whether we call it advanced or not.
1. K. Marx, The Constitution of the French Republic Adopted November 4 1848, Notes to the People, London, no. 7 (June 1851); cited in Hal Draper, Karl Marx's Theory of Revolution, Vol. I (1977), Monthly Review Press, p. 316.
2. Ralph Miliband, The State in Capitalist Society, Weidenfeld & Nicolson, (1969), (now in Quartet paperback); Nicos Poulantzas, Political Power and Social Classes, NLB, (1973); Fascism and Dictatorship (1974); Classes in Contemporary Capitalism (1975); Crisis of the Dictatorships (1976). Miliband and Poulantzas debated in various issues of the New Left Review.
3. John Lea, The State of Society, International Socialism, old series, no 41, Dec.–Jan, 1969; Simon Clarke, Marxism, Sociology and Poulantzas's Theory of the State, Capital and Class, 2, summer 1977.
4. Isaac Balbus, Modern capitalism and the state, Monthly Review, May 1971.
5. John Lea, op. cit.
6. For instance, Simon Clarke, Althusser's Marxism (scandalously still unpublished mimeo, but available from him at Dept. of Sociology, University of Warwick); E.P. Thompson, The Poverty of Theory and other essays, Merlin 1978.
7. Lucio Coletti, From Rousseau to Lenin, NLB (1972).
8. See Colin Barker, Muscular Reformism (review of Marxism and Politics), International Socialism, old series, 102 (October 1977).
9. Nicos Poulantzas, State, Power, Socialism (hereafter SPS), New Left Books, (1978), £7.50. Poulantzas also gave a very interesting interview to Henri Weber of the French section of the Fourth International; this first appeared in Critique Communiste 16, (June 1977), then appeared in translation in International and was reprinted in the US journal Socialist Review, no. 38 (March–April 1978), the source from which I quote (hereafter Interview).
10. SPS, p. 256.
12. ibid., p. 257.
13. Interview, p. 20.
14. Well, almost. There is the following small interchange between Poulantzas and Weber, in which neither is very illuminating:
Poulantzas: Do you believe in pluralism?
Weber: Of course. We believe in it and we practice it.
Poulantzas: But for your opponents as well?
Weber: Certainly. Even for the bourgeois parties, it's there in writing.
Poulantzas: Aha. Even for the bourgeois parties. Now, not to be too naive, there are things one has to say, because we fear for ourselves as well.
Weber: Of course.
Poulantzas: It's all very well to say so, but I want to know what forms of institutional guarantee there would be – they are always secondary, of course, but they matter … (Interview, p. 23)
We leave to some other occasion discussion of the FI's written guarantees to bourgeois parties!
15. SPS, pp. 264–5.
16. ibid., p. 260.
17. ibid., p. 256. Note that Poulantzas, eager to claim the restless spirit of Karl Marx for his ideas, is very cavalier with the old revolutionary, as he is with Engels, who actually suggested that if people wanted to understand 'the dictatorship of the proletariat' they should look at the Paris Commune. It seems that if there was a 'signpost' in Marx and Engels, it did not exactly point in Poulantzas's direction, but directly against him.
18. SPS, p. 124.
19. ibid., p. 125.
20. e.g. Norman Geras, The Legacy of Rosa Luxemburg, NLB 1976, p.187. Others of her criticisms of the Bolsheviks in her pamphlets were, in any case, quite misplaced: cf. Tony Cliff, Rosa Luxemburg, IS 1969, ch. 7.
21. cf. Tony Cliff, op. cit., p. 70.
22. Norman Geras, op. cit., p. 144.
23. ibid., p. 126.
24. I draw here on Tony Cliff, Lenin, Vol. 3, Pluto (1978).
25. SPS, p. 253; Interview, p. 21.
26. In the Interview, Henri Weber offers this argument, but lets Poulantzas get away with murder in reply. Poulantzas refuses to accept the historical explanation of the inward defeat of the Russian Revolution, citing as additional examples of lack of democracy the revolutions in China, Cuba, Cambodia. Weber does not mention – how could he? – that the working class played no independent role in these revolutions, that they never were any kind of socialist revolution. Has the FI come to this pass, that it cannot defend the soviet idea against reformism?
27. Interview, p. 22.
28. Poulantzas defends the secret ballot, as surprisingly does Henri Weber. (Interview, p. 25)
29. e.g. SPS, p. l04, on the 'individualisation of the body-politic as an ensemble of identical monads separated from the State.'
30. cf. e.g. Franz Neumann, The change in the function of law in modem society, The Democratic and the Authoritarian State, Free Press (1957); Poulantzas himself also notes this, e.g. SPS, p. 172.
31. Interview, p. 20.
32. SPS, p.262, emphasis in original.
33. Poulantzas's argument involves overthrowing Marx's opinion, from the Paris Commune onwards, that the working class cannot simply lay hold of the ready-made machinery of the state and wield it for its own purposes. He treats 'representative democracy' as if it were a class-neutral form of eternal relevance. Yet he also contradicts himself: e.g. '... political domination is itself inscribed in the institutional materiality of the State… state power (that of the bourgeoisie in the capitalist state) is written into this materiality' (SPS, p. 14); he notes that the capitalist state embodies the separation of manual and intellectual labour, but fudges the issue in relation to Parliament – 'It is equally clear that a number of institutions of so-called indirect democracy (political parties, parliament, etc.) in which the relationship between State and masses is expressed, themselves depend on the same mechanism' (SPS p. 56) – leaving open the possibility that 'some' institutions are somehow different, but without specifying which, or how.
34. Interview, p. 31.
35. e.g. his chapter on The Weakening of the State, SPS, pp. 241ff.
36. ibid., pp. 206–7.
37. Interview, p. 20
38. SPS, p.255. Cf. also ibid., p. 261.
39. ibid., p. 264.
40. ibid., p. 255.
41. ibid., p. 260.
42. Interview, pp. 13–4.
43. SPS, p. 198.
44. ibid., p. 197
45. ibid., p.197.
46. Inter alia, cf. Nicolai Bukharin, The Economics of the Transformation Period, Bergman (1971).
47. SPS, p. 197.
48. ibid., p. 265.